Although the quality of Lithium-Ion Cells is improving tremendously over the years, the prediction of the lifetime of lithium-ion batteries is still a topic of interest. Economically the lifetime estimation is needed for guarantee issues and mitigating cost risks, when designing electric vehicles and storage applications. For electric vehicles driven by private consumers, it is expected, that those vehicles will be parked or charged most of their lifetime rather than driven. Therefore, besides cycling effects, also calendric ageing has to be understood. A lot of testing data can be found in literature with well established chemistries, but pretty much no data is available for high-energy density chemistries.
This research will present a calendric ageing test matrix on high energy 18650 cells with Nickel-Cobalt-Aluminum (NCA) cathode and a Silicon-Graphite (Si/C) mixed anode. The cell under test is produced by Samsung with a nominal capacity of 3.4 Ah. More than 60 cells were stored at five different Voltage-Levels (State-Of-Charge-Levels) at four different elevated temperatures (35°C to 50°C). The cells were kept constantly at the defined potential (floating test) using Hameg Power Supplies (Rohde&Schwarz). The capacity of the cells were measured every two month by running a “Check-Up” routine at 23°C using an electric cycler from Digatron. Resistance increase of the cells were evaluated using pulses. The test data is now available over a test duration of 2.5 years and the test is still running.
The research results discuss the measured effects. The expected calendric ageing effects are clearly visible for voltages above the 50% SOC-Limit. Temperature influence can be identified and as expected higher temperatures lead to increased capacity decay. The results of the differential voltage analysis (DVA) show that there also seems to be certain amount of active material loss, which is typically related to cycling ageing. This was not expected for a calendric lifetime test and this gives rise to the question how much influence the Check-Up routines have on the measured data. This point is discussed when evaluating the data.